Objective: It was the aim of this study to compare differences in disease pattern, patient characteristics and survival in patient cohorts treated during different decades. Methods: We conducted a retrospective analysis of all patients with brain metastases from colorectal cancer treated between 1983 and June 2008 in Northern Norway. The patients were assigned to 3 different groups, based on the decade of treatment. Results: The time interval between first cancer diagnosis and brain metastases has significantly increased over time. The use of chemotherapy before development of brain metastases has also increased. Only few patients did not harbour extracranial metastases. Chemotherapy after diagnosis of brain metastases has been used exclusively in the present decade, but in only 3 patients. Combined surgical resection or radiosurgery plus whole-brain radiotherapy has increasingly been utilized, but whole-brain radiotherapy alone remained the cornerstone. Neither survival from first cancer diagnosis nor from brain metastasis treatment has improved significantly; however, with up to 17 patients, the groups were small. Three factors were significantly associated with better survival: good performance status, limited number of brain metastases (1 vs. 2–3 vs. 4 or more) and absence of extracranial metastases. The prognostic impact of the recursive partitioning analysis classes was confirmed, while the new graded prognostic assessment index performed less well. Conclusions: Median survival was maximum 6 months in all decades, despite the increasing use of more aggressive treatment. As most patients harbour extracranial metastases that threaten their lives, systemic treatment might theoretically play a role in the management of these patients, but more data need to be collected to confirm the clinical impact of this approach.